Slashdot Mirror
Google Spends $1 Million For Throttling Detection
foamrat writes "Google has awarded $1 million to Georgia Tech researchers so that they can develop simple tools to detect Internet throttling, government censorship, and other 'transparency' problems."
a suite of Web-based, Internet-scale measurement tools that any user around the world could access for free
So, what happens if the Web-based suite is throttled or censored?
The easier to detect, the harder it is for ISP's to keep such practices out of the spotlight.
Really appreciate that google consistently places them in the proper position of an infrastructure provider, setting up their monetization to be supported by open, fair access.
I don't trust "intent," but I do trust a business that is set up to maximize profit when things are best for the "little guy."
Their APIs are a joy to work with, too.
If Country = "Canada"
Then Print "Yes, you're are being throttled!"
Elseif Country = "China"
Then Print "Yes, you're being censored, I hope you can read English or this will be really confusing!"
.
.
.
PROFIT!
.
.
.
End
DAMN! Throttled again... That's Google's project right there. Hiring 1000's of first posters and measuring the delay!
I don't trust Google's altruism further than I can throw it(though it is likely that the founders are better than the outside CEO in that regard...); but it is somewhat convenient that Google's business model and business interests happen to be fairly strongly aligned with most of the internet's virtues.
It cannot be ignored that they want to data-mine you to dump ad impressions down your consumer gullet; but this does mean that they view anybody else trying to do so as "competition" rather than "our bestest buddies in profit"(as with ISPs and Phorm/NSA for instance).
Similarly, their desire to operate communications services without hindrance is not identical to an ideological stance in favor of freedom; but the end results are substantially closer than are those of entities that wish to hinder communications services in order to raise prices, or prevent "piracy", or the like.
I don't trust them; but I would say that their self-interest aligns atypically well with much(though certainly not all) of what would want from the ideal internet. I do trust them to follow their self interest. I would certainly prefer the internet according to some genuinely freedom-focused entity; but those tend to be penniless ragtag optimists, rather than corporate behemoths....
The original gatech study showed not only bufferbloat, but enormous variation of base latencies in the first mile for different brands of cable modem as well as for different kinds of DSL and wireless technologies.
Slides: http://www.caida.org/workshops/isma/1102/slides/aims1102_ssundaresan.pdf
Some commentary: http://gettys.wordpress.com/2011/02/17/caida-workshop/
I look forward to the followup!
What's funny about it? He's basically saying that Google (mostly) does things that he likes, just not for the reasons he'd like them to. It's a perfectly reasonable position to take.
I'm curious to know what about my post marks it as "anti-corporate". As best I can tell, I started from a wholly orthodox account of how corporations operate and worked from there.
Corporations are complex systems designed to serve the (typically financial) interests of their owners. Within the limits imposed by the principal-agent problem, this is supposed to mean that the people who run the corporation serve the interests of the shareholders, by some combination of upping stock prices or issuing dividends.
Because of that, trying to infer the behavior of a corporation from the human motives of the guys at the top is typically going to be a bad model: one should instead expect that the corporation will act in the interests of its shareholders.
I was under the impression that that was pretty much the standard model of corporate behavior. If that counts as "anti-corporate"(rather than, say, somebody actually challenging the notion of 'limited liability investments', or even just asserting that having mercilessly self-interested entities around is a bad idea...), then what kind of bowing and scraping would I need to engage in to be "pro corporate"?